3.6GHz. It is smack back within the spectrum frequencies that telcos and telco authorities worldwide are suggesting is the best to deliver forthcoming 5G technologies.
So, the ACMA has been working on what it needs to do to make 3.6GHz the pioneer band for 5G, and what it means for the industry and existing users of the band.
In October 2016, the ACMA issued a discussion paper entitled “Future use of 1.5GHz and 3.6GHz bands” as an initial investigation, and received 72 submissions to the consultation, and one supplementary submission.
Then, in June this year, the ACMA issued a consultation paper to move to the preliminary planning stage, seeking more industry comment, dubbed “Future approach to the 3.6GHz band”, with a consultation date closing 4 August.
As part of the consultation process, a “spectrum tune-up” event was held on 12 July in the ACMA Sydney, Melbourne and Canberra offices, all linked up via video-conference, providing the opportunity for direct feedback from relevant stakeholders
David Goggin and Mark McGregor from ACMA were present, with the first ACMA speaker giving an overview into the June discussion paper, including proposed options and the reasoning behind the ACMA’s currently preferred option.
We then had Telstra and Vodafone separately give their mobile telco industry perspectives on future 3.6GHz use.
This was followed by a presentation from WISP AU, the Wireless ISP Association of Australia, featuring its secretary, Lynda Summers, and its president, Mike Parnell of Shoalhaven Internet, delivering the perspectives of the incumbent fixed wireless broadband users of the 3.6GHz space.
After this came a presentation from Bob Horton of the Communications Alliance satellite services working group, representing the incumbent satellite industry licensees of the 3.6GHz group.
This was followed by a panel discussion and concluding remarks.
So, what did everyone say? Here's just some of what was discussed:
The ACMA presentation looked at the strong indications globally that the 3300 to 3800MHz band was being targeted for initial deployments of 5G internationally, which the proposed 3.6GHz band fitted right into.
However, with operators desiring at least 100MHz of spectrum in the frequency for themselves, there presented a problem - there is only 125MHz of spectrum available in the band, which will likely see demand exceeding supply in some areas, with existing planning and licensing arrangement not optimal for areas where demand exists.
There were also existing regional users of the 3.6Ghz band, by earth station satellite dish operators, and by regional wireless Internet service providers.
Options exist to get satellite providers to relocate earth stations to other areas, even though this would be very expensive, although quiet zones around existing stations could be set up.
Meanwhile, sharing models of the spectrum were also possible, but how to determine who the top tier user of the spectrum is? Any second tier user would have to use spectrum opportunistically, but lower tier users would not be guaranteed protection from interference.
The ACMA did have a preferred option - 3c in its paper - which included an option for a 7 year reallocation transition, which operators like Telstra, earth station providers and WISP-AU members could work with, but much preferred periods like 10 to 15 years for transition, so investments made could be realised with profits and service provided to existing users.
Naturally, trying to move out existing historical users of the band would be a messy process, even if it this would be the bureaucratically preferred option.
Thus, we had Telstra’s perspective, as it is keen to get moving with trialling 5g technologies in preparation for rolling out 5G networks.
The benefits of 5G are immense, after all. Faster mobile broadband, new services, speeds up to 10x faster than 4G today in excess of 10Gbps speeds.
There are opporunities for network slicing, partitioning, virtual networks with in a network, useful for public safety, or connected vehicles, for example, along with very low latency of a few milliseconds, for remote surgery, real-time control of the smart grid.
There’s also massive scale connected IoT devices, which 5G will handle vastly better than 4G is already doing that today.
However, Telstra also operates earth stations, so any changes to 3.6GHz spectrum for its assets here would also be affected, but as an operator very keen on 5G, it wants to move quickly.
After all, Telstra is conducting trials in the Commonwealth Games timeframe next year, although stressed those trials actually had nothing to do with the Commonwealth Games as had been reported.
There was also the issue of SK Telecom to conduct 5G trials for the Winter Olympics as part of the 3GPP 5G Phase 1 trials, in Q3 2018.
Telstra will also be hosting the 3GPP plenary in Australia in that timeframe.
Q4 2018 would see AT&T commence its 5G deployment, as noted at the Wells Fargo 5G forum on 22 June 2017.
Q4 2019 would see WRC-19 bands about 24GHz explored.
Q4 2019 would see 3GPP go to phase 2, which will add massive IoT capability.
Then in Q2/Q3, the Japan Summer Olympics would see 5G tested, after which more operators in various countries should be starting their own 5G rollouts.
Telstra said early access to 3.6GHz spectrum was critical for Australia to be a 5G player, and needs access to 2018 to keep with global pace.
It supported ACMA’s prioritisation to 3.6GHz, supports ACMA’s preferred option to auction all of the spectrum in metro and regional areas, with existing licensees being protected for an extended period of time, giving new and existing licensees adequate certainly.
Telstra also noted it wears two hats, with earth stations and seven point-to-point links using 3.6GHz, and needs appropriate arrangements to manage that impact.
Telstra proposed that satellite operators be given 15 years to transition, and 10 years for fixed wireless access and point-to-point operators, but that seven years “is also workable”, and that there “may be some merit in new designation quiet zones for the relocation of earth station facilities, with returning about 3700MHz also to be considered, as this would be a lot more cost effective that moving earth stations!"
Next came Vodafone, which is a already doing a lot of work globally, having done early trials getting latency down to 2.8ms, even though actual 5G would have sub 1ms latency speeds.
Vodafone spoke of driverless cars using 5G, remote surgery and more, which without realtime feedback can’t work.
Vodafone said 5G would likely require TDD rather than FDD, and wanted up to 100MHz of TDD spectrum for itself, while noting there was only 125MHz available, and which would be insufficient for the long term needs of the industry.
Stating suburban Australia was the key challenge, with Australia on the leading edge of mobile broadband, 3.6GHz was the sweet spot for coverage and was internationally aligned for scale, cost and roaming.
Vodafone didn’t need to deploy in 2018, but wanted access in 2018, and was wary of deploying technology that was too early, and didn’t want to regret having waited 6 or 12 months more before deployment to get better equipment.
Vodafone also recognised it needed to collaborate with regional ISPs and was developing models to do so.
WISP AU – Wireless Internet Service Provider Association of Australia
Next came Lynda Summers, secretary of WISP AU, and Mike Parnell, president of WISP AU, who explained the issues that concern their members.
Ten years ago, Summers noted WISP was lobbying ACMA for spectrum, and was offered the largely unused 3.6GHz band to affordably and reliably offer service to regional customers – customers who weren’t getting coverage from traditional telcos.
These regional players have been the innovators in regional Internet provision, and were now concerned they were being handed the rough end of the stick in proposals today, with 40 WISP AU members providing services to over 200,000 customers across regional Australia.
WISP-AU wants the productivity commission, other areas of government and more to come together to figure out how to preserve the industry, save a lot of re-investment, share the spectrum and more.
Parnell noted the major challenges facing the WISP AU community and said he felt a heavy burden in representing the community, having acquired 3.6GHz a decade ago, and now potentially being pushed out by major players – and that WISP AU would be working on its submissions for the consultation paper, and noted that with over 200,000 regional Australians using WISP AU member services, that “we are not small.”
I interviewed Parnell after the event, and you can see his thoughts below.
COMMUNICATIONS ALLIANCE – SATELLITE WORKING GROUP
Dr Bob Horton from the Communications Alliance satellite working group was also extremely vocal in addressing concerns that using the 3.6GHz space would be problematic for earth station operators, and that it wasn’t simple or affordable to just up and move to more quiet locations.
He talked about US and European work already done on sharing 3.6GHz spectrum among users without catastrophic excising of incumbents, and advised ACMA to study that work in detail to help resolve its spectrum sharing issues here in Australia.
Dr Horton said a lot more, which presumably will be detailed in his submission paper.
In the panel discussion, all of the above topics were spoken of again, with the various industry representatives talking to each other.
In amongst a lot of detail that will be undoubtedly covered in the various submission papers due by 4 August, there was a question put to ACMA over whether a sharing approach was considered.
The answer from the ACMA representative was that “while we considered it, we couldn’t come up with a model that the incumbents and WISP-AU wanted".
ACMA noted it was “very keen on sharing approaches – the engineering team was heavily involved and considered this in detail, but we couldn’t come up with a sharing model that delivered what was necessary in this case. I’d love to hear it – sharing and dynamic spectrum allocation is the future.”
Parnell noted that the model should “be created by a group of experts in different fields. In the engineers stumble in areas outside engineering, that’s because it’s not their field. The economic, the political issues — national security, sovereign power — that’s not a sphere for engineers”, and wanted various players to come together to sort this out, not just relying on ACMA engineers to get this right.
He said: “we cannot shortcut spectrum sharing because engineers in ACMA couldn’t work out a policy.”
Dr Horton said ACMA could “look for some ideas on sharing policies — report EEC 254 — about operational guidelines for spectrum sharing — 3600 to 3800 MHz — that’d be a good starting point.”
So, despite a lot of concern, it looks hopeful that AMCA, telcos, earth station operators, point to point operators and importantly, WISP AU members, will find a way to share the spectrum, and have enough time to move to other frequencies without hopefully going broke first, but it will require a lot of work for everyone’s mutual satisfaction.
The submissions for this consultation paper will be very interesting to read, with the fate of hard working regional farmers, miners and citizens at stake.
In the meantime, the road to 5G continues, and no matter what happens, we’ll be getting it within the next half decade.
Finally, I offered to video record the entire meeting, which would have been great for all players to be able to refer to should they have wanted to, but sadly did not ask for permission in time to do so to get acceptance from all presenters, so this was not permitted.
Next time I will ask for permission much earlier to see if video recording the event may be possible at that time for those reading this that would have been interested in seeing it.